Synthesis of an Oxazoline Analogue of Apratoxin A - Organic Letters

Aug 27, 2003 - Evolved Diversification of a Modular Natural Product Pathway: Apratoxins F and G, Two Cytotoxic Cyclic Depsipeptides from a Palmyra ...
1 downloads 0 Views 79KB Size
ORGANIC LETTERS

Synthesis of an Oxazoline Analogue of Apratoxin A

2003 Vol. 5, No. 19 3503-3506

Bin Zou, Jingjun Wei, Guorong Cai, and Dawei Ma* State Key Laboratory of Bioorganic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, 354 Fenglin Lu, Shanghai 200032, China [email protected] Received July 18, 2003

ABSTRACT

Michael addition of Me2Cu(CN)Li2 to r,β-unsaturated lactone 7 derived from β-hydroxyl ketone 5 provides lactone 8, which is converted to alcohol 11 using Oppolzer’s methodology as the key step. Connection of 11 with the L-proline moiety and subsequent installation of an oxazoline ring affords 16, which is coupled with tripeptide 21; subsequent macrocyclization then furnishes 4, an oxazoline analogue of apratoxin A.

Apratoxin A (1, Figure 1) is a cyclodepsipeptide bearing both peptide and polyketide moieties isolated from the marine cyanobacterium Lyngbya majuscula by Moore and coworkers.1 One year later, two other analogues, apratoxins B and C (2 and 3, Figure 1) were discovered by further organism collections and isolations.2 Through in vitro studies on cytotoxicity against several human tumor cell lines, apratoxin A was found to have IC50 values ranging from 0.36 to 0.52 nM. However, in vivo antitumor investigation indicated that 1 was poorly tolerated in mice mainly due to lack of selectivity for different cell lines.1 Structure-activity relationship (SAR) studies on this compound might provide a cure for this problem. In fact, the different cytotoxicity pattern displayed by 1-3 has already demonstrated that their cytotoxicity is highly dependent on their primary structures. It was reported that 3 possesses almost identical IC50 values for in vitro cytotoxicity in comparison with 1, but 2 is significantly less cytotoxic than 1.2 To initiate SAR studies (1) Leusch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J.; Corbett, T. H. J. Am. Chem. Soc. 2001, 123, 5418-5423. (2) Leusch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J. Bioorg. Med. Chem. Lett. 2002, 10, 1973-1978. 10.1021/ol035332y CCC: $25.00 Published on Web 08/27/2003

© 2003 American Chemical Society

we decided to explore an efficient synthesis of apratoxin A and its analogues. Recently, Forsyth and Chen disclosed an

Figure 1. Structures of apratoxins A-C and an oxazoline analogue of apratoxin A.

elegant total synthesis of apratoxin A3 based on their newly developed methodology for preparing thiazoline,4 which prompted us to report here our result on the synthesis of 4, an oxazoline analogue of apratoxin A. Obviously, this compound may still have potent cytotoxicity activity as apratoxin A, but its total synthesis may be simpler than that of 1 because there exist more convenient methods to form oxazolines in comparison with thiazolines.5,6 Structurally, 4 contains a novel 3,7-dihydroxy-2,5,8,8tetramethylnonanoic acid (Dtena) unit, which connects with a modified serine moiety (moSer) by an oxazoline ring, and a tetrapeptide unit possessing a high degree of methylation at the L-proline site. We planned to synthesize 4 in a convergent manner as illustrated in Figure 1, that is, peptide formation between the O-Me-Tyr and moSer sites and macrocyclization at the N-Me-IIe-Pro site. The synthesis of enantiopure 3,7-dihydroxy-2,5,8,8-tetramethylnonanoic acid (Dtena) precursor 11 is outlined in Scheme 1. β-Hydroxyl ketone 5 (>99% ee), an aldol reaction product of trimethylacetaldehyde with acetone catalyzed by D-proline,7 was employed as the starting material. After protection of 5 with TBSCl, the ketone moiety was reduced

with NaBH4 and the resultant alcohol was subjected to elimination through its mesylate to afford allyl ether 6a. It is notable that we initially attempted asymmetric allylboration of trimethylacetaldehyde according to Brown’s procedure in order to obtain allyl alcohol 6b.8 However, an inconvenient workup operation (it was found that the desired product was difficult to separate from the resultant isopinocampheol by either distillation or column chromatography) made us give up this direct synthesis. Next, ozonolysis of 6a provided an aldehyde, which was condensed with LiCH2CO2Et at -78 °C followed by cyclization and elimination to give the R,βunsaturated lactone 7. Methylcupration with Me2Cu(CN)Li2 served as an excellent method9 for installing the C37 methyl group in a highly diastereoselective fashion to provide 8 in 94% yield as a single diastereomer as detected by 1H NMR. After LAH reduction of the lactone 8 to produce a diol and then protection of it with acetic chloride, selective deprotection with K2CO3 was carried out in methanol. Subsequent Dess-Martin oxidation of the resulting primary alcohol delivered aldehyde 9. For the assembly of the C33-C35 unit in the target molecule, Oppolzer’s methodology10 for preparing anti-diols from bornanesultam-derived boryl enolates appeared to be particularly attractive. Toward this end, treatment of Npropionylsultam with Et2BOTf resulted in a (Z)-enolate, which was reacted with the aldehyde 9 under the action of TiCl4 to afford “anti”-aldol 10 in 90% yield as the only detectable diastereomer. It is noteworthy that 8 equiv of TiCl4 was necessary in this case to ensure the high yield. Finally, removal of both the chiral auxiliary and the acyl group by LAH reduction gave a triol, which was protected with DMP to furnish 11 in 67% yield. With the alcohol 11 in hand, our next task was to connect it with the L-proline unit on the right side followed by subsequent installation of an oxazoline ring bearing an R,βunsaturated ester side chain on the left side. As outlined in Scheme 2, esterification of 11 with N-Fmoc-L-proline using Yamaguchi’s procedure11 worked well to produce 12 in 90% yield. Liberation of the diol moiety in 12 with TsOH in methanol followed by selective oxidation of the primary alcohol using hindered chloro oxammonium salt generated from TEMPO/NaClO afforded an aldehyde, which was further oxidized with NaClO2 to provide acid 13.12 In a parallel procedure, R,β-unsaturated ester 14, prepared by a Wittig olefination reaction of (R)-Garner aldehyde and the corresponding ylide, was treated with 3:1 trifluoroacetic acid and water to give liberated amine. Coupling of this amine with the acid 13 mediated with HATU and DIPEA (diisopropylethylamine) furnished 15 in 90% yield. The oxazoline

(3) Chen, J.; Forsyth, C. J. J. Am. Chem. Soc. 2003, 125, 8734-8735. (4) Chen, J.; Forsyth, C. J. Org. Lett. 2003, 5, 1281-1283. (5) For reviews, see: (a) Wipf, P. Chem. ReV. 1995, 95, 2115-2134. (b) Wipf, P.; Venkatraman S. Synlett 1997, 1-10. (6) For recent leading references on the preparation of thiazoline and oxazoline rings, see: (a) McKeever, B.; Pattenden, G. Tetrahedron 2003, 59, 2701-2712. (b) Yokokawa, F.; Sameshima, H.; Katagiri, D.; Aoyama, T.; Shioiri, T. Tetrahedron 2002, 58, 9445-9458. (c) Kedrowski, B. L.; Heathcock, C. H. Heterocycles 2002, 58, 601-634. (d) Boden, C. D. J.; Norley, M.; Pattenden, G. J. Chem. Soc., Perkin Trans. 1 2000, 883-888. (e) Boden, C. D. J.; Norley, M.; Pattenden, G. J. Chem. Soc., Perkin Trans. 1 2000, 875-882.

(7) (a) List, B. Synlett 2001, 1675-1685. (b) List, B.; Pojarliev, P.; Castello, C. Org. Lett. 2001, 3, 573-575. (8) (a) Racherla, U. S.; Brown, H. C. J. Org. Chem. 1991, 56, 401401. (b) Jadhav, P. K.; Bhat, K. S.; Perumal, P. T.; Brown, H. C. J. Org. Chem. 1986, 51, 432-439. (9) Pirkle, W. H.; Adams, P. E. J. Org. Chem. 1980, 45, 4117-4121. (10) Oppolzer, W.; Lienard, P. Tetrahedron Lett. 1993, 34, 4321-4324. (11) (a) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull. Chem. Soc. Jpn. 1979, 52, 1989-1993. (b) Hikota, M.; Sakurai, Y.; Horita, K.; Yonemitsu, O. Tetrahedron Lett. 1990, 31, 6367-6370. (12) Anelli, P. L.; Biffi, C.; Montanari, F.; Quici, S. J. Org. Chem. 1987, 52, 2559-2562.

Scheme

1a

a Reagents and conditions: (i) TBSCl/imidazole, DMF, rt; (ii) NaBH4, MeOH, 0 °C; (iii) MsCl/Et3N, CH2Cl2, rt; (iv) t-BuOK, toluene, reflux; (v) O3/Me2S; (vi) LiCH2CO2Et, THF, -78 °C; (vii) 40% HF, MeCN, rt; (viii) MsCl/Et3N, CH2Cl2, 0 °C to rt; (ix) Me2(CuCN)Li2, ether, -78 °C; (x) LAH, THF, reflux; (xi) AcCl/ Py, CH2Cl2, 0 °C, then K2CO3, MeOH; (xii) Dess-Martin oxidation; (xiii) N-propionylsultam/Et2BOTf/DIPEA, CH2Cl2, -15 °C, then 9, TiCl4, -78 °C; (xiv) LAH, THF, reflux; (xv) DMP/PPTs.

3504

Org. Lett., Vol. 5, No. 19, 2003

Scheme 2 a

Scheme 3 a

a Reagents and conditions: (i) 2,4,6-trichlorobezoyl chloride, DIPEA, benzene then 11, DMAP, rt; (ii) TsOH, MeOH, rt. (iii) TEMPO, NaClO, aq NaHCO3, CH2Cl2, 0 °C; (iv) NaH2PO4/ NaClO2, t-BuOH, H2O, 2-methylbutene; (v) 14/TFA/H2O, then 13/ HATU/DIPEA, CH2Cl2, rt. (vi) DAST, CH2Cl2, -78 °C; (vii) Pd(PPh3)4, N-methylaniline, THF, rt.

ring formation was achieved through exposure of 15 to 2.5 equiv of DAST in methylene chloride at -78 °C.13 Subsequent cleavage of allyl ester catalyzed by Pd(PPh3)4 under the assistance of N-methylaniline resulted in acid 16 in 70% yield over two steps.14 It is notable that the utilization of N-methylaniline as the base was essential for this step because other bases employed such as morpholine were found to cause the simultaneous cleavage of the NR-Fmoc protecting group during the above deprotection reaction. For the synthesis of the highly methylated tripeptide part, 2-bromo-1-ethyl pyridinium tetrafluoroborate (BEP), an efficient coupling reagent for hindered peptide synthesis developed by Xu and Li, was used for peptide formation.15 As shown in Scheme 3, deprotection of 17 with diethylamine in acetonitrile followed by amide formation with N-FmocO-methyl-L-tyrosine 18 under the assistance of BEP gave a dipeptide, which was subjected to Pd/C-catalyzed hydrogenolysis to afford acid 19 in 55% yield over three steps. Cleavage of the Fmoc group in 20 and subsequent connection of the liberated amine with 19 mediated with BEP resulted in tripeptide 21. Next, treatment of 21 with diethylamine in acetonitrile produced a free amine, which was coupled with (13) Lafargue, P.; Guenot, P.; Lellouche, J.-P. Heterocycles 1995, 41, 947-957. (14) Ciommer, M.; Kunz, H. Synlett 1991, 593-595. (15) Li, P.; Xu, J.-C. Tetrahedron 2000, 56, 8119-8131. Org. Lett., Vol. 5, No. 19, 2003

a Reagents and conditions: (i) Et NH, MeCN, rt; (ii) 18/BEP/ 2 DIPEA, CH2Cl2, rt; (iii) Pd/C/H2, EtOAc; (iv) 20/Et2NH, MeCN, then 19/BEP/DIPEA, CH2Cl2, rt; (v) Et2NH, MeCN, then 16/ HATU/DIPEA; (vi) TBAF, THF; (vii) HATU/DIPEA, CH2Cl2 (0.002 M), rt.

the acid 16 to afford 22 in 59% yield. The stage was now set for the crucial macrocyclization. We were gratified to observe that after removal of the Fmoc and TMSE protecting groups with TBAF in one pot, the target molecule 416 was obtained in 45% yield by treatment the resultant amino acid with HATU/DIPEA in a diluted methylene chloride solution. (16) Selected data for 4: [R]20D -117.5 (c 0.2, CHCl3); 1H NMR (500 MHz, CDCl3) δ 7.15 (d, J ) 8.7 Hz, 2H), 6.80 (d, J ) 8.2 Hz, 2H), 6.21 (d, J ) 9.3 Hz, 1H), 6.01 (d, J ) 9.3 Hz 1H), 5.25 (d, J ) 11.5 Hz, 1H), 5.05 (td, J ) 10.2, 5.6 Hz, 1H), 4.97 (d, J ) 11.6 Hz, 1H), 4.81 (m, 1H), 4.70 (d, J ) 10.6 Hz, 1H), 4.38 (m, 1H), 4.20 (m, 1H), 3.78 (s, 3H), 3.703.55 (m, 3H), 3.28 (br q, J ) 6.7 Hz, 1H), 3.15 (d, J ) 3.8 Hz, 1H), 3.10 (d, J ) 11.7 Hz, 1H), 2.85 (dd, J ) 12.7, 4.6 Hz, 1H), 2.81 (s, 3H), 2.75 (s, 3H), 2.65 (d, J ) 13.5 Hz, 1H), 2.33 (m, 1H), 2.25 (m, 1H), 2.15 (m, 1H), 2.05 (m, 1H), 1.92 (s, 3H), 1.90-1.75 (m, 3H), 1.45 (m, 1H), 1.31 (m, 1H), 1.26 (m, 1H), 1.25 (s, 3H), 1.10 (m, 1H), 1.07 (d, J ) 6.9 Hz, 3H), 1.01 (t, J ) 7.0 Hz, 3H), 0.99 (d, J ) 6.4 Hz, 3H), 0.96 (m, 1H), 0.95 (d, J ) 7.6 Hz, 3H), 0.87 (s, 9H); ESI-MS m/z 824 (M + H)+; HRMS calcd for C45H70N5O9 (M + H)+ 824.5168, found 824.5160. 3505

In conclusion, we have described here an efficient route to synthesize an oxazoline analogue of apratoxin A. Notable elements include a highly diastereoselective assembly of the Dtena moiety and subsequent installation of the oxazoline ring bearing an R,β-unsaturated ester side chain. In addition, the success in macrocyclization at the N-Me-IIePro site demonstrates again that this site is a suitable macrolactamization site for synthesizing apratoxins and their analogues.3 Extension of these investigations to the total synthesis of apratoxin A and other analogues, as well as the biological evaluation of the synthesized compounds, is being actively pursued in our laboratory and will be reported in due course.

3506

Acknowledgment. The authors are grateful to the Chinese Academy of Sciences, National Natural Science Foundation of China (Grant 20242003), and Science and Technology Commission of Shanghai Municipality (Grant 02JC14032) for their financial support. We thank Professor Tao Ye of The Hong Kong Polytechnic University for helpful discussion and for providing information about their results in this field. Supporting Information Available: Experimental procedures and characterizations for compounds 6a, 7-13, 15, 16, 19, 21, 22, and 4. This material is available free of charge via the Internet at http://pubs.acs.org. OL035332Y

Org. Lett., Vol. 5, No. 19, 2003